BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20261012T210904Z
LOCATION:Summit Pavilion
DTSTART;TZID=America/Los_Angeles:20261021T173000
DTEND;TZID=America/Los_Angeles:20261021T183000
UID:HFESAM_ASPIRE 2026_sess257_POST178@linklings.com
SUMMARY:Systematic Review of Large Language Models (LLM) for Surgical Trai
 ning and Evaluations of LLM Performance
DESCRIPTION:Peiran Liu, Gaoyuan Tu, Aqib Abdullah, Che Liu, Haozhi Chen, a
 nd Jingkun Wang (Purdue University)\n\nSurgical education faces significan
 t challenges in cost, accessibility, and personalized feedback. This syste
 matic review evaluates the application and performance of large language m
 odels (LLMs) in surgical training. We analyzed 34 studies with screening 1
 ,159 publications across four databases. Results show 85% utilized base LL
 Ms, while 15% employed fine-tuned agents. LLMs demonstrated strong utility
  across preclinical knowledge assessment (up to 90% accuracy), intraoperat
 ive decision support (72–76% relevance), and postoperative structured feed
 back. Although domain-specific models outperform general versions, challen
 ges like hallucinations, privacy risks, and algorithmic bias remains. Over
 all, LLMs are a valuable tool to surgical education. Future research shoul
 d prioritize standardized benchmarking, multimodal integration, and rigoro
 us clinical safety evaluations.\n\n
END:VEVENT
END:VCALENDAR